A year ago, everyone thought they knew the best way to treat severe sepsis. A protocol and a campaign encouraged hospitals around the world to follow the same steps in the 6 hours after sepsis diagnosis.
Then 2 large randomized trials—ProCESS in the U.S. and ARISE in Australia—found that patients receiving the early goal-directed therapy (EGDT) protocol weren't any more likely to survive than those who got the participating hospitals' standard sepsis care.
“It would appear that there's no advantage to doing these more rigid protocolized assessments,” summarized John A. Kellum, MD, a ProCESS study author and professor of critical care at the University of Pittsburgh Medical Center. “One doesn't need to implement the package of early goal-directed therapy.”
The shocking news spurred the Surviving Sepsis campaign to put out a press release and make plans for a guideline update later this year. But while some reassessment of sepsis strategies is in order, the trial results showed that current sepsis treatment is generally pretty good, experts agreed.
“Not only were the trials negative—no difference in mortality rates between the treatment and control groups—but [they showed] remarkably low pooled mortality rates, in the range of 18% to 22%,” said James Russell, MD, professor of medicine at the University of British Columbia in Vancouver.
He and other critical care experts had some suggestions on how hospitalists may want to modify care based on the ProCESS and ARISE trials, which were published in the New England Journal of Medicine last year on May 1 and Oct. 16, respectively, as well as on other recent research. Studies have identified a number of issues and uncertainties in sepsis treatment, from whether to use a protocol to how to optimize fluid resuscitation.
“In sepsis, it's been in many ways a good decade, but in many ways a tough decade,” said Allan J. Walkey, MD, assistant professor of medicine at Boston University. “Mortality's been improving markedly in patients with sepsis and septic shock, while a string of trials has shown that nothing particularly is helping them.”
Explaining the new data
There are a number of reasons why recent trials may have failed to find the same benefit from EGDT as the original trial by Emanuel Rivers, MD, and colleagues that was published in the New England Journal of Medicine in 2001 and was the basis for the Surviving Sepsis guidelines.
One possibility is that the initial trial was wrong. “Maybe it never did work,” said Sandra Peake, MD, PhD, lead author of the ARISE trial and an associate professor of medicine at the University of Adelaide in Australia.
It's also possible that the protocol helps only patients like those in the Rivers trial. “The patients in the Rivers trial seem to have been very, very different from the patients in all the other trials looking at sepsis and septic shock. The patients in Rivers' trial had a central venous oxygen saturation in the 40s at baseline,” said Dr. Walkey.
One component of the EGDT protocol is bringing the central venous saturation (ScvO2) up to 70% or higher. “In our trial and also in the ProCESS trial, the mean ScvO2 was about 70%,” said Dr. Peake. “If the ScvO2 is above the target for the entire intervention period, it's hard to imagine how that intervention can actually work.”
There's also the issue of mortality. In the control arm of the Rivers trial, 46.5% of patients died, more than double the rate in the more recent trials. Part of that is due to better, faster diagnosis. “You're now picking up patients that you might have missed before,” said Dr. Kellum.
In addition to improved diagnosis, patients now get more rapid treatment, even if they're in the control arm of a study, experts agreed.
“The culture has changed so much over the last 10 years that even in absence of an established protocol, the patient that comes in to those emergency departments gets rapid identification, early antibiotics, and significant early fluid resuscitation,” said R. Phillip Dellinger, MD, an author of the Surviving Sepsis guidelines and a professor of medicine at Cooper University Health Care in Camden, N.J.
Thus, one explanation for the protocol's failure in research is its success in practice. “The success of the Surviving Sepsis campaign almost predicated that these trials would end up negative, because it was going to be hard to go back to the bad old days, where patients were identified late and languished and didn't get adequately resuscitated,” said Dr. Russell.
No one wants to go back to that standard of care. All of the experts agree on the benefits of the earliest components of treatment—effectively, the 3-hour bundle for sepsis care. “What one needs to do is ensure that patients are promptly evaluated, fluid resuscitated when necessary, given antibiotics,” said Dr. Kellum.
Issues of debate
Experts disagree, however, on whether a strict protocol is needed to provide this care. “I'm not necessarily a big one on protocols,” said Dr. Peake. “The problem with protocols is that a) they don't apply to the patient you've got in front of you and b) sometimes [because of them] people don't give that individualized care which patients need.”
Dr. Dellinger disagreed, arguing that the sepsis protocol provides needed guidance for many hospitals. Although the trials found usual care to be just as good, those control arms were at high-quality facilities where expert clinicians knew they were under research scrutiny.
“At community hospitals that don't have the clinical trial in sepsis mentality, these things are not going to be done as quickly. If you look at the ProCESS and ARISE trials, the time to antibiotics in these patients was legend. You would get a gold star,” he said.
There might be some facilities where more rigid guidance is needed, conceded Dr. Kellum. “Austere environments, rural settings, places where there wasn't necessarily the same level of clinician assessment—would you in those situations be better off having a gadget do the monitoring because you don't necessarily have the trained clinicians to do so? The studies don't speak to that,” he said.
The gadget in question is a central line for monitoring central venous pressure. The 6-hour bundle of sepsis care calls for using one to guide fluid resuscitation, a component of the protocol that the ProCESS and ARISE trials did not provide to their control groups, with no apparent impact.
That's the biggest practice change indicated by the trial results, according to Dr. Russell. “I certainly don't recommend use of that central venous catheter any more, and I used to,” he said. “People can safely and very effectively manage septic shock without necessarily feeling compelled to put in the central catheter.”
Recommendations on catheter use will probably be changed in the next update of the guidelines, said Dr. Dellinger, although he still thinks that some hospitals might provide better sepsis care with the definitive guidance provided by the devices. “It may be that in institutions with a different culture than what it was in these clinical trials, that a [central venous] catheter may flag that patient as a patient that needs more attention, that needs more fluids,” he said.
That argument ignores the costs of catheter placement, countered Dr. Peake. “If you had an [unnecessary] therapy that isn't harmful, that's one thing, but if the therapy that isn't harmful actually costs more than another therapy, then I don't think you can advocate doing that therapy,” she said. “We had a significant number of patients who were admitted to the ICU, not for clinical reasons, but only for delivery of the early goal-directed therapy.”
In addition to financial costs of extra protocol components, there's potential opportunity cost. “While physicians are busy putting in specialized catheters and attending to specialized monitoring and following complicated protocols, that time could be better served by doing other things which are important for patient care,” said Dr. Kellum. “Some patients are going to require more resuscitation and some patients are going to require less, and the physician's role is to determine that.”
There is an option between adhering strictly to the existing protocol and making all one's own decisions, suggested Dr. Walkey. “Having something in place might be helpful for places that have very little experience in caring for sepsis, but that protocol might not have to be early goal-directed therapy. Maybe it could be the other protocol in the ProCESS trial that didn't require a central line and ScvO2 monitoring,” he said.
Whether they want protocols or not, even the experts would like some more guidance about fluid resuscitation.
Underresuscitation has been the historic problem in sepsis treatment, but it's possible that patients could eventually get too much of a good thing. “There is some risk that the pendulum will swing too far, and we'll be giving too much fluid to patients,” said Dr. Kellum. “Intravenous fluids are not benign ancillary therapies that can be treated with a different set of cautions than other forms of medication.”
Timing of fluid delivery appears to be an important consideration. Everyone knows early treatment is good, but giving a lot of fluid later may actually be harmful, according to Rahul Kashyap, MD, assistant professor of medicine at Mayo Clinic in Rochester, Minn.
“Our retrospective study suggests that people who [survived] got more fluid in the first 3 hours,” he said. “If you are delayed out of that window, then you are doing more harm to the patient if you give more fluid.”
The choice of fluid for resuscitation may also affect patient outcomes. “Which fluid? It's crazy that it hasn't been answered yet,” said Dr. Walkey. The ProCESS and ARISE trials provided no conclusions on this issue.
“Because [ProCESS] was conducted in the U.S., the vast, vast majority of fluid was 0.9% saline,” said Dr. Kellum. “The ARISE trial would have used different fluids, because in Australia, there's a tendency to use different kinds of fluids than in America. And they got the same results.”
Other recent studies have suggested that all fluids are not the same, though. Albumin appeared to help patients with septic shock, although not severe sepsis generally, in an Italian study published in the New England Journal of Medicine on April 10, 2014. “In that group, the albumin lowered the mortality,” said Dr. Russell. “It doesn't mean people should be using lots of albumin, but I think it will raise the ante. I think we'll see further studies and trials that could lead one day to the suggestion that albumin is preferred in septic shock.”
Lower-chloride solutions may also become more popular for sepsis treatment based on new research. “There was some evidence that normal saline or any high-chloride solutions increase the risk of developing acute kidney injury,” said Dr. Russell. “I would certainly like to see us using more of these non-chloride-containing solutions or very low chloride concentration. They're not much different in price.”
It would be more costly, but the problem's solution might be the development of new solutions, according to Dr. Kellum. “Would it be better to give patients a multielectrolyte solution that more closely mimics their plasma volume?” he said. “Is there a need to develop better fluids? Is there a need to specifically personalize the fluid prescription? That's what I feel.”
Greater personalization of treatment is probably the future of sepsis treatment generally, other experts agreed. “They are looking for early biomarkers to define sepsis to see who requires aggressive treatment,” said Dr. Kashyap. Biomarkers may also be the new outcome in sepsis research, since ProCESS and ARISE showed that it's difficult for an intervention to make a difference in mortality rate, Dr. Russell noted.
He'd also like researchers and clinicians to remember that although early sepsis treatment is important, so is what you do later. “We need to focus on the survivors,” he said, describing his study that followed some patients for 10 years after sepsis treatment. “One episode of septic shock was the equivalent of taking 12 years off your … life expectancy.”
That finding should inspire hospitalists to action. “It means we need a lot of research and probably a lot of rehab,” Dr. Russell said. “Instead of just sending them home and saying, ‘Good luck.’”